Cell-Specific Calcineurin Inhibition by a Modified Cyclosporin

Abstract

The immunosuppressive drug cyclosporin A (CsA) is widely used to suppress graft rejection following transplantation in humans. Its cellular effects are mediated by a complex of CsA and its intracellular, soluble receptor cyclophilin (Cyp).1 The Cyp-CsA complex is a specific inhibitor of the protein phosphatase calcineurin (Cn).2 Consequently, CsA has also been used to illuminate the cellular functions of Cn. For example, its use revealed that, in T cells, Cn mediates the nuclear translocation and activation of NFAT,3 a transcription factor essential for T cell activation. As a therapeutic agent, CsA has several undesirable side effects including hypertension, CNS toxicity, and nephrotoxicity. These side effects result from CsAmediated inhibition of Cn in cells outside the immune system.4 We now report the synthesis of R-cyclopentylsarcosine11CsA (CsA*), a rationally modified CsA that does not bind to Cyp and consequently does not inhibit Cn in cells. We also describe a modified Cyp with compensatory mutations in its CsA-binding pocket (F113G, C115M, S99T; Cyp*) that promote high affinity complexation with CsA*. The resulting Cyp*CsA* complex presents a composite surface that binds Cn with high affinity. As a result, Cn inhibition by CsA* is restored in cells expressing Cyp*. These results provide a method to control the inhibition of Cn spatially and temporally in animals by targeting the expression of Cyp* to specific cells and tissues. Modifying receptors or ligands by adding even a few atoms can abolish their binding, often changing the function of the molecules involved. When these loss-of-interaction substituents are introduced into proteins by mutation5 and small molecules by synthesis,6 information about the cellular function of interacting molecules can be obtained. To achieve our goal of cellspecific Cn inhibition, it was essential that the modified receptor-ligand complex have a composite surface suited for Cn binding, since altered Cyp-CsA complexes can be envisioned that will no longer bind Cn (lower right schematic in Figure 1). We designed a variant of CsA having two additional methylenes (CH2) attached to the two methyl groups of the MeVal11 side chain, resulting in the cyclopentyl side chain of CsA*. The crystal structure of the CypA-CsA complex7 shows residue 11 of CsA directly contacting Cyp, binding in a deep hydrophobic pocket in the active site of Cyp (Figure 2A). The additional atoms were expected to reduce the binding of CsA* to Cyp significantly, presumably through steric interactions